Power Analysis of Data Center Workloads

Undergraduate Project

Investigators:  Christopher Martinez, Assistant Professor
                          William Putnam, B.S. in Computer Engineering and Computer Science

Funding:  Summer Undergraduate Research Fellowship 2013

Project  Description: Our society is becoming more dependent on cloud technology and services. The cloud has changed the computer needs for the end-user and the provider. The end-user now requires better battery performance than computational horsepower. The content provider is now required to provide 24-hour reliable and fast service. The result has been that data centers are consuming vast amount of energy worldwide.

In this research project, we are examining the power consumption of different CPUs on common data center workloads. The data that we collect on ARM CPUs and Intel CPUs will enable society to understand the power being consumed by the new cloud services and provide insight on how hardware/software designers can reduce power consumption.

The project is wide in scope and requires the need for both computer engineering and computer science at both the graduate and undergraduate level.